phys. stat. sol. (b) 241, No. 3, R5–R7 (2004) / DOI 10.1002/pssb.200309026© 2004 WILEY-VCH Verlag GmbH & Co. KGaA, Weinhei...
R6 O. Maksimov et al.: Temperature dependence of the energy gap of MgxZnyCd1–x–ySe alloy© 2004 WILEY-VCH Verlag GmbH & Co....
phys. stat. sol. (b) 241, No. 3 (2004) / R7© 2004 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheimphysicastatussol...
Upcoming SlideShare
Loading in …5

Temperature dependence of the energy gap of MgxZnyCd1–x–ySe alloy


Published on

phys. stat. sol. (b) 241, No. 3, R5–R7 (2004)

Published in: Technology, Business
  • Be the first to comment

  • Be the first to like this

No Downloads
Total Views
On Slideshare
From Embeds
Number of Embeds
Embeds 0
No embeds

No notes for slide

Temperature dependence of the energy gap of MgxZnyCd1–x–ySe alloy

  1. 1. phys. stat. sol. (b) 241, No. 3, R5–R7 (2004) / DOI 10.1002/pssb.200309026© 2004 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheimphysicastatussolidi⋅RapidResearchNoteRapid Research NoteTemperature dependence of the energy gap ofMgxZnyCd1–x–ySe alloyO. Maksimov*, 1, Martin Muñoz2, W. H. Wang1, M. C. Tamargo2, and N. Samarth11Department of Physics, Pennsylvania State University, University Park, PA 16802, USA2Department of Chemistry, City College of New York, New York, NY 10031, USAReceived 22 December 2003, revised 19 January 2004, accepted 20 January 2004Published online 26 January 2004PACS 78.30.Fs, 78.40.Fy, 78.55.EtThe temperature dependence of the energy gap of MgxZnyCd1–x–ySe (x ∼ 0.5) and Zn0.5Cd0.5Se alloys wasdetermined from reflectivity measurements at temperatures between 4.2 and 300 K. The data were fittedto a Bose–Einstein type relationship and temperature dependence of band gaps was compared. A similartemperature dependence of band gap was observed indicating that quantum confinement in theMgxZnyCd1–x–ySe/Zn0.5Cd0.5Se quantum wells did not depend on temperature.© 2004 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim1 Introduction The quaternary MgxZnyCd1–x–ySe alloy is a promising material for the fabrication of fullcolor light emitting and laser diodes (LEDs and LDs). MgxZnyCd1–x–ySe based red, green, and blue LEDs[1, 2], current injected LDs (operating at liquid nitrogen temperature) [3], and optically pumped lasers[4] were demonstrated. These light emitting devices are usually grown by molecular beam epitaxy(MBE) on InP substrates (metal organic vapor-phase epitaxy of MgxZnyCd1–x–ySe was also reported [5])and consist of a Zn0.5Cd0.5Se quantum well (QW) layer (Eg ~ 2.15 eV) and MgxZnyCd1–x–ySe barrier layers(Eg ~ 3.0 eV, x ~ 0.5). However, in spite of a significant progress achieved, no current injected LDs op-erating at room temperature were reported so far. Also, a significant increase in the threshold pumpingintensity with the rise in temperature was observed for the optically pumped lasers due to the leakage ofelectrons from the well to the barriers, indicating insufficient quantum confinement [6]. It was suggestedthat the quantum confinement can be improved either by increasing the band gap of the barrier layers orby using artificial “multi-quantum barriers”. To optimize the electronic structure of the active area, it isimportant to know the dependence of band gap energy on temperature. Temperature band gap variationof Zn0.5Cd0.5Se was studied by photoluminescence (PL) spectroscopy [7]. Studies of MgxZnyCd1–x–ySeshowed that the PL emission was dominated by the localized excitons and did not follow the band gap[8]. Thus, the temperature band gap variation of MgxZnyCd1–x–ySe is unknown.We report the use of PL and reflectivity spectroscopies to study the temperature band gap variation ofMgxZnyCd1–x–ySe (x ~ 0.5). The temperature dependence of the band gap energy is determined from re-flectivity spectra and modelled by a Bose–Einstein type Eq. [9]. It is observed that temperature depend-ence of the MgxZnyCd1–x–ySe band gap is similar to that of Zn0.5Cd0.5Se. Thus, quantum confinement doesnot significantly change as a function of temperature in the MgxZnyCd1–x–ySe/Zn0.5Cd0.5Se QWs, simplify-ing the design of the active region of the light emitting devices.2 Experimental details All of the samples were grown by MBE on semi-insulating InP (100) sub-strates under the growth conditions described elsewhere [1]. The composition of the ternary ZnxCd1–xSe*Corresponding author: e-mail:, Phone: +1 814 863 9514, Fax: +1 814 865 3604
  2. 2. R6 O. Maksimov et al.: Temperature dependence of the energy gap of MgxZnyCd1–x–ySe alloy© 2004 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheimphysicastatussolidi⋅RapidResearchNotealloy was calculated from the lattice constant measured by X-ray diffraction, assuming a linear depend-ence of lattice constant on the alloy composition (Vegard’s law). The composition of the quaternaryMgxZnyCd1–x–ySe alloy was determined by the combination of lattice constant and band gap energy data[10]. PL measurements were performed in a liquid He continuous flow cryostat in the temperature rangefrom 4.2 K to 300 K. Excitation was provided by the 325 nm line of a He–Cd laser. The collected PLwas spectrally resolved by a monochromator and detected by a cooled charge coupled device array de-tector. Reflectivity measurements were performed in the same cryostat using a 200 W Hg lamp.3 Results and discussion PL and reflectivity spectra at 4.2 K for the samples grown are shown inFig. 1 by the solid and dotted lines, respectively. In order to elucidate the origin of the emission lines, thedependence of the emission on the excitation laser intensity was studied (not shown). The excitation laserintensity was varied over three orders of magnitude and a linear dependence of the PL emission with aslope near unity was obtained. Also, no shift of the emission energy as a function of the excitation laserdensity was observed. This behavior is consistent with the excitonic origin of the MgxZnyCd1–x–ySe emis-sion [11]. The quenching of Fabry–Perot oscillations in the reflectivity spectrum corresponds to theonset of interband absorption and is used to identify position of the Γ → Γ direct band gap with an uncer-tainty of ~6 meV. An increase in the Stokes shift, given by the difference in the energies of absorptionedge and emission line, from ~10 meV for Zn0.5Cd0.5Se to ~90 meV for Mg0.54Zn0.26Cd0.2Se is evident.A significant Stokes shift observed for MgxZnyCd1–x–ySe indicates that localized exciton emissiondominates the spectra. When studied as a function of temperature, the emission energy did not follow theband gap and exhibited a characteristic ‘S-shaped’ behavior (decrease–increase–decrease) due to theionization of localized excitons to free excitons [8].The temperature dependence of the direct band gap was studied by reflectivity measurements. Thedata are shown by symbols in Fig. 2. As the temperature increases, the critical point exhibits a red shiftdue to the decrease in the band gap energy caused by the electron–phonon interactions and lattice dila-tion. The data are fitted to the equation proposed by Viña et al. (solid lines) [9]:EBE (T) = E(0) – 2aB/(exp (θ/T) – 1) , (1)2.0 2.5 3.0 3.5Egx= 0.45EgEgEgx = 0.54x= 0.51x= 0T = 4.2 KIntensity(A.U.)Energy (eV)10 1002. 0.45x= 0.54x= 0.51x= 0Energy(eV)Temperature (K)Fig. 1 Photoluminescence (solid lines) and reflec-tivity (dotted lines) spectra for MgxZnyCd1–x–ySeepilayers.Fig. 2 Temperature dependence of the MgxZnyCd1–x–ySeband gap energy. Symbols are experimental data. Thedotted lines are theoretical fits to the Bose–Einstein typeequation.
  3. 3. phys. stat. sol. (b) 241, No. 3 (2004) / R7© 2004 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheimphysicastatussolidi⋅RapidResearchNoteTable 1 Parameters obtained by modelling the temperature dependence of band gap energy to the Bose–Einstein type equation.alloy E(0) (eV) aB (meV) θ (K) dE/dT (meV/K)Zn0.5Cd0.5Se 2.170 22 ± 2 146 ± 13 0.30Mg0.45Zn0.3Cd0.25Se 2.918 25 ± 2 167 ± 5 0.30Mg0.51Zn0.28Cd0.21Se 3.014 26 ± 2 184 ± 8 0.28Mg0.54Zn0.27Cd0.19Se 3.051 25 ± 2 169 ± 8 0.31where E(0) is the energy of the transition at 0 K, aB represents the strength of the electron–average pho-non interaction, and θ corresponds to the average phonon temperature. The band gap shift is proportionalto the sum of the Bose–Einstein statistical factors for phonon emission and absorption. Fitting parametersare summarized in Table 1. From comparison, it is evident that electron–phonon interaction constant andaverage phonon temperatures of MgxZnyCd1–x–ySe are either very similar or slightly higher than that ofZn0.5Cd0.5Se. (The difference is close to the experimental error.)At the high temperature limit, Eq. (1) reduces toEBE(T) = E(0) – 2aBT/θ (2)and the temperature band gap variation (dE/dT) is proportional to –2aB/θ (Table 1). The temperaturedependence of the MgxZnyCd1–x–ySe band gap is similar to that of Zn0.5Cd0.5Se. This is unlike to a morewidely studied ZnSxSe1–x–ZnyCd1–ySe material system, where the band gap of ZnSxSe1–x reduces fasterthan that of ZnyCd1–ySe. Study of ZnSxSe1–x/ZnyCd1–ySe QW structures demonstrated a decrease of thequantum confinement at higher temperatures that induced a leakage of carriers from the well to the barri-ers quenching the emission [12]. The thermal effect is expected to be smaller for MgxZnyCd1–x–ySe/Zn0.5Cd0.5Se QW structures due to the stability of quantum confinement.4 Conclusion In conclusion, MgxZnyCd1–x–ySe alloys (0 < x < 0.5) were grown and their band gap wasstudied as a function of temperature. The temperature dependence of the band gap was measured and thedata were fitted to the Bose–Einstein type equation. The band gap variation of MgxZnyCd1–x–ySe (x ~ 0.5)was similar to that of Zn0.5Cd0.5Se. Therefore, temperature-independent quantum confinement is expectedin MgxZnyCd1–x–ySe/Zn0.5Cd0.5Se QW structures.References[1] M. C. Tamargo, W. Lin, S. P. Guo, Y. Y. Luo, and Y. C. Chen, J. Cryst. Growth 214, 1058 (2000).[2] W. Faschinger and J. Nürnberger, Appl. Phys. Lett. 77, 187 (2000).[3] S. B. Che, I. Nomura, A. Kikuchi, and K. Kishino, Appl. Phys. Lett. 81, 972 (2002).[4] L. Zeng, B. X. Yang, A. Cavus, W. Lin, Y. Y. Guo, M. C. Tamargo, Y. Guo, Y. C. Chen, Appl. Phys. Lett. 72,3136 (1998).[5] M. Straßburg, O. Schulz, U. W. Pohl, D. Bimberg, D. Litvinov, D. Gerthsen, M. Schmidbauer, and P. Schäfer,J. Cryst. Growth 221, 416 (2000).[6] X. Zhou, M. Muñoz, M. C. Tamargo, and Y. C. Chen, J. Appl. Phys. 95, 7 (2004).[7] O. Maksimov, W. H. Wang, N. Samarth, M. Munoz, and M. C. Tamargo, Solid State Commun. 128, 461(2003).[8] X. Zhou, Y. Gu, I. L. Kuskovsky, G. F. Neumark, L. Zeng, and M. C. Tamargo, J. Appl. Phys. 94, 7136(2003).[9] L. Viña, S. Logothetidis, and M. Cardona, Phys. Rev. B 30, 1979 (1984).[10] O. Maksimov, S. P. Guo, M. C. Tamargo, F. C. Peiris, and J. K. Furdyna, J. Appl. Phys. 89, 2202 (2001).[11] T. Schmidt, K. Lischka, and W. Zulehner, Phys. Rev. B 45, 8989 (1992).[12] I. Lo, K. H. Lee, L. W. Tu, J. K. Tsai, W. C. Mitchel, R. C. Tu, and Y. K. Su, Solid State Commun. 120, 155(2001).